X-linked severe combined immunodeficiency (XSCID) is characterized by profound defects in both T-cell and B-cell immunity. A naturally occurring animal model for human XSCID exists in a breeding colony of dogs, which have clinical, genetic, immunologic, and pathologic features indistinguishable from those in human XSCID patients. It is likely that the human and canine diseases are caused by mutations at a homologous gene locus, but the specific biochemical and genetic defects are not known in either species. The immediate goal of this proposal is to isolate the canine XSCID gene, with the long-term objectives of demonstrating that defects in the homologous gene in humans cause human XSCID, characterizing the XSCID gene product and its role in immune function, and using the XSCID dog model to develop approaches to somatic cell gene therapy for immunodeficiencies. This will be accomplished using a novel positional cloning (reverse genetics) approach that takes advantages of both human and canine disease features and genetics. The human XSCID locus resides on the proximal short arm of the X chromosome (Xq13). Boundaries for the XSCID locus in the dog will be delineated by linkage analysis using our large pedigree of XSCID dogs and probes adapted from cloned loci in human Xq13. Candidate canine XSCID cDNA clones will be isolated by screening canine lymphoid cDNA libraries with yeast artificial chromosomes (YACs) that contain very large fragments (~250 kbp) of human Xq13 chromosomal DNA. These candidate clones will be evaluated for expression in lymphoid RNA and for DNA sequence alterations correlated with the presence of the XSCID phenotype in the dog colony. XSCID cDNA candidates will subsequently be analyzed for expression and function after transfection or retroviral infection into cultured lymphoid cells from XSCID male puppies. Ultimately, the role of the XSCID gene in lymphoid cell development will be studied in normal and XSCID dogs. The approach outlined takes advantage of complementary features of the human and canine systems. Contributed by the human system are: a large number of independent XSCID mutations, well ordered and characterized Xq13 marker loci, and X-chromosome YAC libraries. Hybridization of YACs across species will minimize the high background caused by repetitive sequences. The canine model offers: a multi-generation pedigree with very large sibships, the ability to produce three generations of controlled matings in two years, and the availability of cells and tissues from all stages of development. Moreover, the canine XSCID model is ideal for performing the controlled experiments, many of which are not possible in human patients, necessary to evaluate approaches to immunotherapy and gene therapy for the treatment of immunodeficiencies.